The heat transfer coefficient between the alloys and cooling water is affected by a lot of factors and hard to measure, a new method was investigated with a self-designed system ultilizing SP-15 high-frequency inductive heating unit. Based on measured temperature curves and Fourier heat transfer model, quantitative correlation between heat transfer coefficient and temperature was obtained by inverse algorithm method of iterative simulation and automatic optimization. The results showed that in submerged water-cooling process, the heat transfer coefficient reached to a peak value at the beginning and then decreased with increasing temperature. A decrease of cooling water temperature increased the peak value of the heat transfer coefficient, but did not change temperature range of the peak value from 200°C to 225°C . The heat transfer coefficient was mainly dependent of interfacial temperature between the Al-Cu alloys and the cooling water.The temperatures range from 200°C to 225°C gave the highest heat flux transfer.